I am trying to wrap my head around FM modulation, and have setup some experiments in GNU Radio to help, but they have only raised more questions.
I am trying to achieve a peak deviation of 2400Hz using a baseband signal of 1200Hz. In my flowgraph [here](
http://imgur.com/5pzSnsq), and [here](
http://imgur.com/uBDP3kF), you can see that I've set the fm block with a sensitivity of (2*pi*2400)/(48000), and inversely for the gain of the quadrature demodulation block.
Using a baseband of 15Hz, I also get results that I expect, (a signal oscillating from ~-2400 to 2400Hz](
http://imgur.com/VzG6bxx).
Where my results differ from what I expect, is approximately above 100Hz. Using an baseband of 1200Hz, I no longer get a signal sweeping from -ive to +ive, I get [a constant frequency of 1200Hz on the waterfall](
http://imgur.com/BqGsRtK), and aliased multiples.
I believe that my sample frequency of 48000 is sufficient as my mf is 2400, and baseband is 1200.
BW is (2 * (1200 + 2400)) = 7200Hz
Nyquist is 2*BW = 14400Hz
So, what am I doing wrong? Thanks in advance!